magnitude
Open source inference server that runs the best local models for your hardware, plugged into the agent you already use. Works with Pi, OpenCode, Hermes, OpenClaw, Codex, Claude Code, Oh My Pi, and Cline.
Understand Magnitude's SDK bundle plugin version check inheritance. Learn how the @magnitudedev/sdk package ensures automatic RPC version validation without custom plugin logic.
How the OpenAPI Effect Layer Exposes RPC Endpoints in magnitudedev/magnitudeDiscover how the OpenAPI Effect layer in magnitudedev/magnitude compiles OpenAPI specs into typed RPC endpoints. Learn about operation descriptors and manifests for SDK client access.
How the Inference Benchmark Package Measures tok/s in MagnitudeDiscover how the inference benchmark package in magnitudedev/magnitude measures tok/s by dividing tokens by time, reporting median throughput for prefill and decode phases.
Design Document Discovery Workflow in Magnitude: A Complete GuideMaster the Magnitude design document discovery workflow. Learn how the bun design-docs CLI links source files to design docs, ensuring architectural integrity before commits.
How the Harness Integration Connects to External Agents in magnitudedev/magnitudeLearn how the harness integration connects to external agents using the HarnessConnection service and HarnessConnectorRegistry in magnitudedev/magnitude. Manage harnesses seamlessly.
How to Use the Effect Language Service to Analyze Files in MagnitudeAnalyze Effect-TS declarations and dependencies with the Effect Language Service CLI tool in magnitudedev/magnitude. Inspect graphs and layers without full compilation.
How the Magnitude SDK Handles Fixed-Endpoint Admission and RecoveryDiscover how the Magnitude SDK manages fixed-endpoint admission and recovery. Learn about its semaphore and JIT-RPC transport for reliable connections.
Branded Types for ID Semantic Attribution in the Magnitude CodebaseDiscover how Magnitude uses Effect-TS branded types for type-safe ID attribution. Learn to create nominal types to prevent accidental mixing across domains.
How the ICN-Engine Scheduler Handles Speculative Decoding in MagnitudeLearn how the ICN-Engine scheduler in magnitudedev/magnitude handles speculative decoding. Discover efficient KV caching and sequence reuse to eliminate redundant prefills and boost performance.
How to Run Magnitude's Test Suite with Vitest and the --bun FlagLearn to run magnitudedev/magnitude's test suite with Vitest and the --bun flag. Execute tests faster using Bun runtime for improved performance and access to Bun globals.
Package Layering Architecture from Clients to ACN in Magnitude: Complete GuideUnderstand magnitudedev/magnitude's package layering architecture from clients to ACN. Explore the strict four-tier hierarchy and how each layer enforces its public API for robust development.
How the AI Provider Abstraction Works in Magnitude: Unifying Heterogeneous Model ProvidersDiscover how Magnitude's AI provider abstraction unifies diverse model services like OpenAI and Anthropic into interchangeable TypeScript components. Swap providers effortlessly without code changes.
Have a question about this repo?
These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:
curl -s "https://instagit.com/install.md" 
Maintain an open-source project? Get it listed too →